Cracked.com is running a Photoshop contest – “The retarded Truth behind the World’s Greatest Mysteries”. Truth be told, my favourite entry is mrlarry’s tame dinosaur moving Stonehenge’s stones in its mouth. But there are a couple of Voynich Manuscript themed entries too…

voynich-sanchez-small
Sanchez’s competition entry
 (quarter size)

The VMs as drawn by a child? Sweet Photoshop technique, but probably not quite ‘retarded’ enough to win. The second entry is a bit like a hi-tech version of David N. Guy’s Voynich “Breakthrough” we saw a few days ago:-

voynich-danthegreat-small
Dan The Great’s competition entry
 (quarter size)

Yes, it does indeed read “With great haste he removed her underpinings. The hour of fornication was nigh.” Hmmm… perhaps the smart money’s still on the dinosaur. 🙂

Incidentally, the Voynichese (from f15v) appropriated here just happens to be the two lines where I think the author used a space transposition cipher (i.e. moving spaces around within the ciphertext) to hide the repetitive “or or or” and “or or or or” by writing them as “oror or” and “or or oro r” respectively [Curse, p.160]. 🙂  It’s also a bit odd that the first letter of the second line is rendered as EVA “l” – in Jorge Stolfi’s transcription, it’s weirdo #138, more like a malformed EVA “s”. Oh well.

Incidentally, John Stojko translated this same passage as “Why are you measuring the measure? The measure is the same.  Even after Great One, the bones will be broken. I am telling you. Relic should believe me. ” I’ll leave it to you to decide whether (Dan’s) Babel Fish’s “With great haste…” is better or worse, I couldn’t possibly judge such a beauty contest. 🙂

Spurred on by a blog comment left this morning, I wondered whether the Dorabella cipher might actually (because of the symmetry of its cipherbet shapes) be some kind of rotating pigpen cipher, where you rotate each of the positions around after each letter. This would be a bit like a “poor man’s Alberti cipher disk”… just the sort of thing a self-taught cipher hacker such as Elgar might devise.

And so, I decided (being a programmer) to code it up. Of course, it didn’t appear to solve it (these things never do), but I thought I’d post my C code here anyway. Enjoy!


#include <stdlib.h>
#include <stdio.h>
char dorabella[] =
 "BLTACEIARWUNISNFNNELLHSYWYDUO"
 "INIEYARQATNNTEDMINUNEHOMSYRRYUO"
 "TOEHOTSHGDOTNEHMOSALDOEADYA";
#define ELEMENTS(N) (sizeof(N) / sizeof(*(N)))
#define DORABELLA_SIZE (ELEMENTS(dorabella) - 1) // trim the trailing zero!
void dorabella_encipher(int c, int *row, int *column)
{
 if (c >= 'V')
  c--;
 if (c >= 'J')
  c--;
 c -= 'A';  // c now equals 0..23
 *column = c % 8;
 *row    = c / 8;
}
int dorabella_decipher(int row, int column)
{
 int c = (row * 8 ) + column;  // space inserted to stop smiley being inserted!
 c += 'A';
 if (c > 'I')
  c++;
 if (c > 'U')
  c++;  // c now equals 'A'..'Z'
 return c;
}
int main(int argc, char **argv)
{
 int i, j, c;
 int row, column;
 int step_size = 1;
 if (argc > 1)
  step_size = atoi(argv[1]);
 for (i=0; i<8; i++)
 {
  printf("C%d: ", i);
  for (j=0; j<DORABELLA_SIZE; j++)
  {
   dorabella_encipher(dorabella[j], &row, &column);
   column += i + (j / step_size);
   while (column < 0)
    column += 8;
   column %= 8;
   c = dorabella_decipher(row, column);
   printf("%c", c);
  }
  printf("\n");
 }
 for (i=0; i<3; i++)
 {
  printf("R%d: ", i);
  for (j=0; j<DORABELLA_SIZE; j++)
  {
   dorabella_encipher(dorabella[j], &row, &column);
   row += i + (j / step_size);
   while (row < 0)
    row += 3;
   row %= 3;
   c = dorabella_decipher(row, column);
   printf("%c", c);
  }
  printf("\n");
 }
 return 0;
}

[Here’s a guest posting from my friend, the well-known Voynich contrarian Glen Claston. Though he originally intended it as a comment to my recent post on Voynich Manuscript Quire 8 [Q8], it actually deserved a whole post to itself. I’ve lightly edited it to house style, and added a couple of pictures.]

Nick asked me to look at the blog, and though I don’t plan to be a regular poster, he’s going on about things that matter a great deal, so we need to examine them very carefully.

The [“ij”] mark at the bottom of f57v is in line with two other erased marks on folios, as well as erased symbols in Q1 and Q2 that Nick and I discussed some time ago. The original author apparently used a symbol system instead of a standard system, and much of his work has been removed. Quirization and foliation are not the work of the original author. I plan to publish this in a book entitled “The Curse of the Curse of the Voynich”. 🙂

f42r-folio-number-closeup
Voynich Manuscript f42r folio number

Rene Zandbergen also brings up an interesting observation about f42r, that the crystals appear to be on top of the foliation. Yes, I’m certain that is the case, but I don’t reach the same conclusion that Rene does in this regard. In the image above, the foliation clearly overlaps the drawing lines and the color pigment, but at the same time, this entire region is a section exposed to water damage, which might explain Rene’s observation of the crystals overlapping the burned-in ink. IF the pigment includes mineral salts as many commonly do, this would explain Rene’s observations, as they would have re-crystalized over the existing material. It would require a rather closer inspection to see if this is the case.

If Rene’s observation holds, there is indeed something seriously wrong with the Voynich, since foliation before coloration has a very dramatic implication on known VMs construction, and I for one would have to throw out years of research and start anew, as would many. I recall that I had issues with Rene before on “retouching” because these darker patches fell into areas that were also affected by moisture. As it happens, I do apologize much belatedly to Rene for suggesting that just because some of these didn’t match, his identification of retouching in the astrological section was wrong, when it proved to be spot-on. It was my fault for generalizing, and to say that we all make mistakes is not a good enough excuse, I owe it to myself and to the VMs to be as precise as possible.

[Nick: as far as the paints go, I think the consensus now is more that different paints were added at different times, though I suspect the “light painter” / “heavy painter” binary division may well prove to be far too simplistic – because of the large number of paints present, I can quite conceive that these might have been added by four or five later “heavy painters”.]

As far as the rosettes section [Q14] goes, Nick is suggesting here that Q14 belongs to Q8, and though I wouldn’t exactly state that in the way Nick has suggested, I entirely agree. The rosettes is a part of the astronomical discussion, so it’s not in its right place. A large folio can get ripped out rather easily, and be placed back in the book in random order. It’s an hypothesis, but is it testable?

It turns out that the rosettes contains a record that helps us place some items in order. There are tears in the unused fold of the rosettes, damage beyond what normal foldouts have seen, and these tears hold information. The quire mark only works if the rosettes was bound in this torn seam, and the foliation only works if the binding of this foldout is in its current place. That says that the maker of the quire marks was not the maker of the original foliation, that the foliation was a product of at least one successive binding. There are two distinguishable hands in the foliation, two successive bindings after the quire mark binder. Two inks that I can identify in the foliation bindings, and places where quire marks were added that weren’t there originally.

rosette-folding

[Nick: GC is proposing that the nine rosettes fold-out f86 was originally attached to the rest of the manuscript along the (now badly damaged) crease highlighted green (above), rather than along the crease highlighted blue. The shape of the whole codex is highlighted in red.]

It’s a complicated picture that needs a degree of clarification, but there is no way around the idea that the manuscript went through at least three bindings. The big question is – does any of these bindings reflect the original order of construction? The answer is a resounding NO.

Again I go to the rosettes for history, and I need only look on the back of the rosettes to see that the discussion includes the four seasons and the four winds. All my research into parallel texts says that this is the meterological part of the astronomical discussion, and belongs firmly in the astronomical section. This also says that the rosettes on the reverse is a meteorological mappa mundi, and read in that venue of comprehension it’s imagery becomes meaningful and schematic to other VMs imagery. It helps that one page in the first astronomical section [in Quire 8] exhibits similar damage to that of the original rosettes page, and Nick is right that the pages in this section appear to be inverted.

I’ve been through the whole range of arguments over the years, and I weary of argument that doesn’t move me forward, but this is a discussion that needs to be moved forward on several fronts, and I will follow this discussion with interest.

What I am seeing is that the quire numbers were placed on probably the first binding, but I’ve always been of the opinion that the manuscript existed unbound during much of the author’s life. I now have much more information to back up that idea, and as you know, I was once of the opinion that it was the author that first quirized, which is something I can now disprove in abundance. Order changes and shuffling I can’t comment on, but there is evidence that the author himself made some major changes, and these changes were substantially reflected on the first binding but the manuscript was not in a permanently bound condition when the author left off/died.

Rene for one would understand that in making these decisions, I’m weighing intelligent choice against mishap, and using a set of parallel texts on these subjects to determine which is which. One doesn’t even require these options in viewing the interlacing of herbal-a and herbal-b herbals. One does however, need to know why the herbal-a herbal pages were separated from the herbal-a pharmaceuticals and additional information stuffed in between, much of this in a different script. Herbal-a herbals are congruent with herbal-a herbals in the pharmaceutical section, the latter sometimes drawn on the same bifolio/foldouts as the pharmaceuticals, and as Nick and I discussed recently, there is physical contact information that ties them together in a time-line of construction. These are connected in multiple ways to the same time line, and the intervening information is connected to a separate time line, and the construction is a progressive construction, so could this have been the act of binder and not the author? Could this intelligent construction occur passively, and not actively? There’s an argument in there somewhere.

Grant for a moment that I don’t think the book was bound during the author’s life, and I am certain it was not bound before the drawings/text/ paints were added (it’s damned hard to draw, paint, and write all the way into bound gutters on so many pages – common sense observation, eh?). What’s just as important is when quire marks ceased and foliation began. Dee used bifolio quire markings in his book of 1562, and though page numbering was becoming popular in printed documents by this time, Dee chose not to use it, choosing a manuscript format instead. It’s a generational thing, and I feel that the foliation is at least 17th century. The quirization has a problem with dating as Rene pointed out, that it could be someone older that didn’t use the modern format, or it could have been someone before the modern format became prominent. The rosettes’ gutter damage however, says that there was a good deal of time between the quire marks and the foliation, because they couldn’t possibly have happened at the same time, and the quire marks are apparently a good deal older than the foliation.

Make of this all what you will! — Glen Claston

A while back, I posted about how the name “Voynich Manuscript” implicitly contains two lies – for one, it had been known for centuries prior to Wilfrid Voynich; and for two, you can only properly call something a “manuscript” if it was produced before 1450, which seems not to be the case here (though probably by only a whisker).

Much the same kind of thing goes for any discussion of “quires” – this too is carrying a heap of linguistic baggage which we should at least be aware of. So: what exactly is a quire, then?

Probably the best starting point is to step through the vellum manufacturing process from the start. Once a suitable animal (typically calf, lamb or goat kid) had been killed and its rawhide (literally its “raw hide”) flayed from the carcass, that hide had to be soaked and pounded (to remove the flesh from the “flesh side” and the hair fibres from the “hair side”) and then “tanned” (to prevent it from rotting). The beautiful thing is that soaking a rawhide for several days in just the right kind of noxious goo (this is called “liming” if you use lime, or “bucking” if you use wood ash) transforms its protein structure and thereby turns that hide into something that will (if kept properly) stay stable for decades, if not centuries – leather.

However, to be able to write satisfactorily on it, freshly-limed leather has to be dried on a stretching frame, “scudded” (to remove any remaining hairs), smoothed with a pumice stone, pounced (roughened up again, otherwise it would be too slippery to write on), and whitened. What you end up with is a piece of writable leather – vellum or parchment. Having said that, given that I have seen at least ten different opinions on what should be called “vellum” and what should be called “parchment”, my best advice is simply to call all writable leather “vellum” and move swiftly on.

Of course, because animal skins don’t come in perfectly book-sized and -shaped dimensions, the large piece of vellum you have just produced then has to be folded, cut and trimmed into a more useful size. By doing this in the most obvious way, you would end up with a nested set of bifolios (double pages with a fold down the middle) where each facing pair of folios has a flesh side or a hair side. This isn’t necessarily always the case, but it does seem to be a rule of thumb that is generally followed.

[Incidentally, the plural of folio is folios, of folium is folia, of bifolio is bifolios, and of bifolium is bifolia. Ignore any Internet nonsense that mixes any of these pairs up, as well as the occasionally-seen foolish assertion that the plural of folio is folio’s. *sigh*]

What you are then holding is called an unsewn gathering – a folded set of loose bifolios. If these are then sewn together (the needle holes down the central fold are called the “sewing stations”), they are called a sewn gathering.

But this is not yet a quire. Strictly speaking, a quire is a part of a set of sewn gatherings that have been sequentially ordered (the ordering mark is known as the “quire signature”, typically numbers or letters – ‘i’ / ‘ii’ / ‘iii’, ‘a’ / ‘b’ / ‘c’,  etc) ready to be bound together – until such time as . There are therefore two basic forms of quires: unbound quires (a set of gatherings that have been ordered but not yet bound) and bound quires (a set of gatherings that have been both ordered and bound). You can also sometimes find an unordered set of gatherings bound together: but without quire signatures (which were normally placed on each quires’ back page), these aren’t really quires per se.

Confusingly, many texts also refer to gatherings or quires as “quire signatures”: though this is a good example of metonymy (using a part to stand in for a whole), in this context it only really serves to muddy the waters.

In the case of the enciphered Voynich Manuscript, we have a basic codicological dilemma: even though the bifolios have been sewn and bound together into quires, there seems little doubt that the various people who added the quire numbers (i.e. the quire signature numbers) did so without being able to read the contents – and because the original bifolio grouping, sewing, and binding order has been lost, we don’t know what the original gathering structure was.

enciphered-x
Voynich Manuscript f83v – enciphered “X”?

And so if the Voynich Manuscript turns out to have enciphered quire numbers (for example, given that f83v was probably the back page of a quire, the drawing there might contain an enciphered “X” to denote “Quire #10”), then it originally was an ordered set of quires (though whether unbound quires or bound quires isn’t known) – but if it didn’t, then it was originally an unordered bundle of gatherings. Or it might have been partially ordered (and so be a mixture of quires and gatherings). We just don’t know… yet!

Elmar Vogt’s blog has just thrown one of his interesting Voynich Thoughts into the air (or should I say “the aiiir”?).

Having patiently tabulated all the paragraph stars in Q20 back in 2006 (it’s a kind of Voynich researcher rite of passage, I did much the same thing in 2002), and classified them into “hollow” [-], “dotted” [o] and “massive” [x], Elmar now notes a repetitive pattern: that most star sequences simply alternate between hollow and dotted, i.e. o-o-o-o-o-o-o-.

Yet because the stars on f103r (which was very probably the original first page of the quire) are heavily embellished, he tentatively concludes that “the writer/painter started off somewhat artistically ambitious, but […] got bored and decided to simply take turns between hollow and dotted stars.

Elmar then points out that “the only pages significantly deviating from this pattern are f103r, f104r and f108r“, and (the logical chap he is) suggests that this might have been caused by the bifolios within the quire being scrambled before the folio numbers were added, i.e. that f103 and f104 (currently the first two folios of the quire) could well have originally been immediately followed by f108 (currently the sixth folio of the quire). All very sensible – but this, just as with every other page ordering hack that’s been proposed over the years, would need additional corroboration from physical evidence, such as handwriting continuity, ink continuity, quill bluntness continuity, coding system continuity, contact transfers, vellum cuts and folds, vellum thickness mapping, etc.

In some ways, this gives me a bit of a weird feeling: I spent years working basically alone on precisely this kind of marginal codicology, but now Glen Claston and Elmar Vogt are both on pretty much the same case. Circa 2005, my point of departure was that the folio numbers and the quire numbers were just plain wrong, and that it simply has to be easier to try to decipher the VMs in its original page order than in its present “anagrammed” (or rather “folio-shuffled”) state.

Even so, we’re attempting to reconstruct the original page-order with our eyes nearly shut: we still need someone to take a huge lateral step sideways to amass a sufficient amount of physical evidence to make any significant progress – but unfortunately the painstaking process of debating one marginal blob at a time is all we currently have open to us. Oh well!

Has Robert Teague found a sensational astronomical ‘crib’ into the Voynich Manuscript’s ciphertext? Several Voynicheros have asked me to have a look at his claim: normally, this is researcher code for “I think it’s nonsense but I’d like someone else to say it rather than me, because I quite like the guy“, but let’s see what he has to say…

Certainly, Robert’s best-known previous attempt at understanding Voynichese (Teague numbers) didn’t work out particularly well – as I recall, he used the table on f49v’s margin as a basis for linking glyphs to numbers, much as Robert Brumbaugh did back in the 1970s. However, given that there is a powerful palaeographic argument this table was added roughly a century after the VMs was originally made, this is a hugely unreliable thing to be basing anything substantive upon.

So, what of Teague’s 2009 assault? He starts out (in his “Cracks in the Ice I” document) by pointing out what he thinks are seven fuzzy matches to “Aldebaran” across several pages, and so links seven Voynichese letters with their Latin plaintext equivalents. The obvious problem with this is that this basic fuzzy template can also be matched throughout the entire text: and you’d have to admit that the notion of the whole of the VMs’ text’s being about Aldebaran is somewhat unlikely. But it’s possible, of course.

He moves on (in his “Cracks in the Ice II” document) to finding a secondary crib for the star Alcyone: however, because I’m pretty sure that Giovanni Battista Riccioli first named this in his (1665) “Astronomia Reformata“, this is probably not correct. Robert also suggests some anagrammatic cribs for HOLLAND, POLLAN (“Poland”), and LAPLAND, all of which seem historically anachronistic (for example, Holland wasn’t known by that name until the 17th century). He also makes extensive use of some letter substitutions suggested by Philip Neal, but almost certainly not in a way that Philip himself would feel particularly comfortable with. He finishes up with a suggested translation for the so-called “PM curve word” (EVA ‘oalcheol‘) as ‘COBBLED”: this is done by picking one out of 128 possible permutations and then anagramming the result, a kind of wobbly mid-ground between Brumbaugh and Newbold. Why he chose ‘cobbled’ and not (for example) the rather more august 13th century Anglo-Saxon ‘BOLLOCS’ you’ll have to decide for yourself.

Finally, in his “Cracks in the Ice III“, Robert moves on to try to work out the rest of the alphabet, but runs into trouble with the much-used ‘4o’ token, to which he assigns a rather arbitrary set of 14 possible letter pairs. At the end, Robert proposes a set of six 21-letter mappings, and presents them in a mysterious colour-coded table, which (frankly) doesn’t make a lot of sense to me.

Of course, this is the point in the post where I’m supposed to say something withering, dismissive and ironic to leave readers chortling into their morning cup of coffee: but that’s not even close to what’s going through my mind. Right now, I actually feel a huge sadness that for many people these days this kind of thing is what passes for credible research.

In many ways, the Internet has de-skilled historical research: I can quite imagine that many students would now be able to gain a history degree without ever entering an archive, without learning Latin, and without actually physically engaging with the subject. If you’re only one paltry mouse-click away from a plausible answer, why bother to look any further? Why, then, should we be cross with non-specialist historians who replicate this same behaviour?

What we have here, then, is simply misdirected cryptology built on top of poor history, with substantial similarities to Robert Brumbaugh’s attempts three decades ago. Though Brumbaugh was an extremely able and clever scholar, he nonetheless read the Voynich Manuscript just plain wrong – and this is the “same old same old”.

Here’s a nice piece of 3D art where the model’s face and corset are real, but everything else is rendered. The artist (‘jfrancis‘ from Los Angeles) has also included (in the post immediately following) a description of how he achieved the effect (with PhotoShop and Maxwell). For maximum Cipher Mystery brownie points, he also included some nonsense Voynichese (such as the EVA “Klobal” at top left and bottom right, though I doubt it means anything) around the edges (what do you mean, “I’m the only person looking at the edges”?)

apnea_knives_sfw_v02_200x300

Having said that, his ciphertext is only Voynichese-like (or “Ruggish”, to use the technical term): “b” is a very rare letter, and he hasn’t quite – even though he does use “or” a lot – got its internal word structure nailed (uppercase doesn’t help). Perhaps he ought to play Voynich Scrabble? =:-o

Don’t say I don’t try to broaden your mind. 🙂

PS: here’s some more Voynich writing used in an “enigmatic instrument” you might also like!

Here’s the nice little video for David Byrne’s (2008) song “The People Tree”. It mashes up 1920s collage stylings (such as cloche hats) with a man in a black mask being interviewed while holding a mysterious book. Lots of Voynich-like bits (plants and trees, nymph-like people), but with a bit of a Codex Seraphinianus edge to it. Sure, I’m too big a fan of David Byrne’s music to be completely objective: but I enjoyed this vijjo & hope you do too!

PS: a big hat tip to the Xenophilius blog for picking up on this!

Given the amount of time cipher mystery researchers spend banging their heads against the limits of knowledge (read: brick walls), the fact that these mysteries sometimes invade their dreams should be no great shock. Here’s a fine recent example of a Voynich dream from an anonymous correspondent (and no, it’s not me being coy):-

“I received in the post a 3-inch-thick wodge of eccentric handwritten notes from a mysterious Voynich researcher who had given up after years of work. The notes were very odd and wild and rather disturbing in nature, as if the product of a mental patient or someone very close to the edge, with some pages containing wildly-scribbled bits of english and other writings in uncontrolled or child-like hands. This was interspersed with other pages of interesting diagrams that looked potentially meaningful or useful. Then I came across a page which had a crayoned self-portrait of the author, who turned out to be a famous hoaxer: the pages that followed seemed to detail how the hoax was done, including illustrations showing how a stylised signature and the date 2009 had been hidden inside a page of pure Voynichese in a sort of “join the dots to see the letters” way. After that it got really weird and there were pages detailing the author’s radical Christian beliefs and even some nutty religious T-shirts that were inside the stack of papers.”

This is far from the only one: Robert Firth’s dream of 2nd February 1992 has all the richly hallucinogenic texture of iconology, yet he still manages to see someone sifting salt and pepper as representing ” the consonants and vowels of the Voynich script”, which Robert interprets to mean that “we shall never be able to separate them.” More recently, Gloria Amendola blogged about a 2008 dream (and a poem) inspired by seeing the VMs at first hand.

As for my own Voynich dreams: though I not so long ago blogged about a Belinda Carlisle / Voynich dream, my single most lasting Voynich dream image is of walking around an ecclesiastical basement (something like a monastery), and suddenly finding  the design of the Voynich ‘t’ (with its distinctive symmetrical double-leg gallows) in a high-up pane of stained glass with the late afternoon sun behind it, in a vivid palette of reds, yellows, and black leading. Yes, I do dream in colour. 🙂

Are dreams able to cast useful light on what we are thinking about? Personally, I have woke up many times with insights and inventive solutions to problems that have been on my mind. But you may be surprised to know that in about 1887 a (still very young) William Romaine Newbold wrote an apparently still-cited article on how problem-solvers are sometimes able to devise answers in their dreams. Perhaps there is hope for us all, then – sweet dreams! 😉

I recently bumped into a “Random Freebie” from blogger “Miss Black Pepper”: apropos of nothing much, she posted up a Voynich Manuscript Scrabble Tile Set, composed of nice little images clipped from the Voynich Manuscript (which she happens to rather like).

Of course, this set me wondering what a real Voynich Scrabble would look like. You see, the most appalling thing about this idea is that it is actually possible, in much the sameway that you don’t need to know what an ULVA is (don’t ask, but many people apparently eat it) to use it to score in Scrabble.

For Voynichese, the problem is that the alphabet used for most transcriptions (called EVA Hand 1) is a stroke-based alphabet, designed to capture what the letters look like (rather than what they are). Though elegant in its own linguistic way, this unfortunately led to a number of (frankly) rather odd design decisions, with a number of apparently unitary “glyphs” being split up into overlapping component bits. If you reassemble the most obviously glyph-like stroke groups back into glyphs, you could just about get a moderately workable set of Scrabble tiles. Note the four tall letters in the top-left row (these are called “gallows”) and the four “vowel-like” glyphs in the top-right row (these are transcribed as “a, e, i, o”).

voynich-scrabble-eva
Voynich Scrabble tiles (using the EVA transcription)

Voynich researchers have spent a lot of time working out abstract models for Voynichese, tiny generative grammars that attempt to predict what should (or indeed shouldn’t) be a valid Voynich word: Jorge Stolfi’s “core-mantle-crust” paradigm is one such model. But the awkward fact remains that none of these works massively well, either generating many more words than actually appear or generating far too few words of those that do appear in the ms.

(And anyway, the question remains what kind of vocab could ever be algorithmically generated… but let’s stay on track here).

You could use Voynichese tiles to play one of two basic types of Voynich Scrabble:

  1. You can only play words that actually appear in the Voynich Manuscript
  2. You can only play words as predicted by [Person X]’s generative grammar

(Quick hint: don’t play rule-set #1 against Philip Neal or Glen Claston, they’ll probably thrash the pants off you.)

Another factor that hasn’t been considered enough is the specific language differences that occur within the Voynich Manuscript. Prescott Currier identified ‘A’ and ‘B’ languages: but labels have a different usage pattern too, as do line-initial, paragraph-initial and word-initial letters. And so you might add a special kind of dialect ‘mode’ to your Voynich Scrabble, that only lets you play ‘A’-language words or ‘B’-language words (or label words) at certain times.

However, one problem I have is that I simply don’t trust the so-called ‘Voynich vowels’. For example, Voynichese ‘o’ functions in so many different ways within the text that I cannot easily believe it is a letter per se: rather, I would argue that it is a multi-function ‘trick’ token, that performs different jobs in different contexts. I’m also deeply suspicious about the ‘a’ and ‘i’ letters, particularly when they appear in ‘aiin’ groups: I argued in my book that these are used as a kind of visual framework to encode numbers using the backwards flourish of the “n” (which seems more like a ‘v’). And so I think the ‘vowels’ form a great big hole for linguists to jump willingly into (but that’s another story entirely).

Anyway, given that I have also long suggested that a small set of letter-pairs / letter-groups would form a more usable starting point for cryptanalysis than ‘raw’ EVA, here are the tiles I’d expect the original author to be most comfortable playing with (you’ll notice the absence of free-standing ‘a’, ‘i’ and ‘o’ from the set). Agree or disagree, here they are:-

voynich-scrabble-paired
Voynich Scrabble (using Nick Pelling’s verbose pairs)

Happy Voynich Scrabbling! 🙂

PS: OK, “ulva” is the name of the sea lettuce ulva lactuca, an edible algae / seaweed. Aren’t you glad you don’t play Scrabble competitively? 🙂